Implementation of the Clustering Using Representatives (Cure-sne) Algorithm on Toddler Stunting Data

This study investigates the application of the Clustering Using Representatives (Cure-sne) algorithm to analyze toddler stunting data, focusing on four key variables: age, gender, height, and nutritional status. The Cure algorithm was chosen for its ability to effectively cluster datasets with diverse shapes, sizes, and distributions, and for its robustness against outliers, ensuring reliable results even in the presence of extreme data points. The data underwent thorough preprocessing, including encoding categorical variables like gender and normalizing numerical variables such as age, height, and nutritional status, to ensure comparability across features. Once preprocessed, the Cure algorithm was employed to identify distinct clusters of toddlers, each characterized by specific combinations of the key variables. This clustering allowed for the identification of children at higher risk of stunting, providing critical insights for the development of targeted interventions. The results have significant implications for public health efforts, enabling more precise and impactful strategies aimed at reducing stunting and improving child nutrition and health outcomes on a larger scale. This study underscores the value of advanced clustering techniques in public health research, particularly in addressing complex issues like childhood stunting.